Bet Spread: How 1-to-12 Beats 1-to-8 (and When It Doesn't)

Spread is what turns a count into money. A 1-to-12 ramp captures roughly 35% more EV per hour than 1-to-8 — but the bigger top bet draws backoffs faster. The published numbers, the bet-ramp economics, and where each spread is the right one.

Two Hi-Lo counters sit at adjacent 6D H17 DAS LS 3:2 tables, $25 minimum each. Both have basic strategy locked in, both know the Illustrious 18, both have the same $20,000 bankroll. The first counter runs a 1-to-8 spread: $25 at TC+1 or below, $50 at TC+2, $100 at TC+3, $150 at TC+4, $200 at TC+5+. Her published edge is roughly 0.55% per round actually played, generating around $11/hr at 80 hands/hour. The second counter runs a 1-to-12 spread: $25 baseline, ramping to $300 at TC+5+. His published edge is roughly 0.75% per round, generating around $15/hr at the same pace — about 35% more dollars per hour for the same skill and bankroll. He's also four times more likely to be flat-betted or backed off at any given casino within his first dozen sessions. That's the spread trade-off in one paragraph: more spread, more EV, more heat.

Bet spread is the bridge between the count and the bankroll. The count tells you when the player has an edge; the spread tells you how big a bet to put down when that edge appears. Get the spread right and the count compounds into real hourly EV; get the spread too small and you're a hobbyist with a 0.1% edge; get it too wrong-shaped and you're a backed-off player who spent two months drilling Hi-Lo for nothing. This post is the published edge numbers for the standard spread sizes, the bet-ramp economics that determine which TC gets which unit, the heat trade-offs, and the spread-by-bankroll guide that working APs actually use.

What 'spread' means precisely

Spread is the ratio of your maximum bet to your minimum bet. A 1-to-8 spread means your top bet is 8x your minimum: $25 min / $200 max, or $50 min / $400 max, or $100 min / $800 max. The ratio matters more than the absolute numbers — a 1-to-8 spread at any unit size produces roughly the same EV-as-percent-of-bankroll.

The 'ramp' is the schedule that maps true count to bet size. The standard ramp encoded in lib/strategy/counting.ts as the betSpreadUnits function looks like:

TC ≤ +1: bet 1 unit (the minimum)
TC = +2: bet 2 units
TC = +3: bet 4 units
TC = +4: bet 8 units
TC ≥ +5: bet 12 units (the maximum)

That's a 1-to-12 ramp. The same player on a 1-to-8 ramp caps the top bet at 8 units instead of 12 — TC+4 stays at 8 units, TC+5+ caps at 8 units. A 1-to-16 ramp continues to 16 units at TC+5 and 20 units at TC+6 if the dealer hasn't shuffled by then.

The shape of the ramp matters as much as the top-bet ratio. A Kelly-style ramp (which is what the standard 1-2-4-8-12 schedule approximates) increases the bet size roughly with the true count — bigger edge, bigger bet, more EV captured at the high-edge moments. A linear ramp (1-2-3-4-5) under-captures the high-count EV; an over-aggressive ramp (1-3-9-15-25) over-bets the marginal counts and exposes the player to higher RoR than the bankroll supports.

Published edge numbers by spread

At a 6D H17 DAS LS 3:2 game with Hi-Lo counting and 75% penetration (a typical mid-strip shoe), the published edge by spread runs approximately:

Flat bet (1-to-1): -0.50% (you lose at the house edge — counting alone with no bet variation captures roughly nothing)
1-to-2 spread: -0.40% (still losing; not enough spread to overcome the house edge in negative-count rounds)
1-to-4 spread: -0.10% (breakeven-ish; not a winning game)
1-to-6 spread: +0.30% (positive edge, but thin)
1-to-8 spread: +0.55% (the working-counter floor)
1-to-12 spread: +0.75% (the typical pro spread)
1-to-16 spread: +0.90% (high-EV; high heat)
1-to-20 spread: +1.00% (specialist; very high heat; rare in modern casinos)

The shape: edge climbs sharply from 1-to-4 to 1-to-8 (each ratio doubling adds roughly 0.20-0.25% of edge), then flattens — going from 1-to-12 to 1-to-16 adds only 0.15%, going from 1-to-16 to 1-to-20 adds only 0.10%. Diminishing returns set in fast above 1-to-12 because the top-bet rounds are rare; capturing them at 16x instead of 12x is a small percentage gain across the total hands played.

The 1-to-12 spread captures about 35% more EV than 1-to-8 (0.75% vs 0.55%). That's the headline trade. Most working APs settle into 1-to-12 as the sweet spot.

How the bet ramp economics actually work

Casino edge at TC 0 in a typical 6D H17 game is roughly 0.50% against the player. At TC+1 the player has roughly 0.0% edge (breakeven). At TC+2 the player has roughly +0.50% edge. Each TC point past +1 adds about +0.50% to the player edge in this game — published in detail in the Wong and Schlesinger references.

True counts distribute through a typical 6D shoe roughly as follows (per Snyder's published distribution):

TC ≤ 0: ~55% of rounds
TC = +1: ~18% of rounds
TC = +2: ~12% of rounds
TC = +3: ~7% of rounds
TC = +4: ~4% of rounds
TC ≥ +5: ~4% of rounds

Multiply edge-at-each-TC by frequency-of-each-TC by bet-at-each-TC, and you get the per-round EV the spread captures. Take the 1-to-12 ramp at $25 base:

TC ≤ 0 (55% of rounds, -0.50% edge, $25 bet): -$0.069/round contribution
TC = +1 (18%, 0.00%, $25): $0/round
TC = +2 (12%, +0.50%, $50): +$0.030/round
TC = +3 (7%, +1.00%, $100): +$0.070/round
TC = +4 (4%, +1.50%, $200): +$0.120/round
TC ≥ +5 (4%, +2.00%, $300): +$0.240/round

Sum: about +$0.39/round. At 80 rounds/hour, that's +$31/hour for a $25-base 1-to-12 spread. Cross-check: 0.75% edge × average bet (about $52 across the spread) × 80 hands/hour ≈ $31/hour. The two methods agree.

Take the same calculation for a 1-to-8 spread (TC+5 capped at $200 instead of $300, TC+4 at $200, TC+3 at $100): the TC+5 contribution drops from $0.24/round to $0.16/round. Total per-round EV drops from $0.39 to about $0.31. At 80 rounds/hour, $25/hour vs $31/hour — the 1-to-12 spread is earning about $6/hour more, which lines up with the 0.75% vs 0.55% edge published figures.

The heat trade-off

Casinos watch top bets. A player whose bets oscillate between $25 and $200 is a $175 bet variance; a player whose bets oscillate between $25 and $300 is a $275 variance — about 57% bigger to surveillance's eye. The probability of being flagged grows roughly with the top-bet-to-min-bet ratio and the abruptness of the ramp (jumping from $25 to $300 in one round looks much more like a counter than ramping $25 to $50 to $100 to $300 across four rounds).

Working figures (anecdotal, not formally published, but consistent across the community):

1-to-4 spread: rarely draws attention. Most casinos ignore it.
1-to-8 spread: occasional attention at watchful properties. Most casinos let it ride.
1-to-12 spread: backed off at the most aggressive properties (Strip high-limit rooms, certain off-Strip properties with strong counter-detection) within 2-5 sessions. Most other properties let it ride for 10-30 sessions.
1-to-16 spread: backed off at most properties within 1-3 sessions. Specialist territory.
1-to-20+ spread: backed off within 1-2 sessions almost anywhere. Team-play or one-shot move only.

The spread that gets you backed off captures zero EV per hour going forward. A 1-to-12 spread that gets you flat-betted after three sessions is materially worse than a 1-to-8 spread that runs unobserved for thirty. Match the spread to the property's heat tolerance, not to the published EV chart. Most working APs run multiple spreads — 1-to-12 at low-heat properties (locals casinos, smaller off-Strip), 1-to-8 at watched properties (mid-Strip), 1-to-4 at the most aggressive properties (high-limit rooms where they want to stay welcome). The bet ramp at /advanced/bet-spread shows the per-TC unit allocation; matching it to the local property is the working AP's daily decision.

Bankroll matters too

Spread doesn't exist in isolation from bankroll. A 1-to-12 spread requires a bankroll that can absorb the variance the top bet generates. Standard rule (Schlesinger via the sibling post at /blog/risk-of-ruin-bankroll-formula): bankroll should be at least 200x your top bet for a 1% RoR target. So:

$25 base × 12 = $300 top bet × 200 = $60,000 recommended bankroll for the published 1-to-12 spread to run at <1% RoR
$25 base × 8 = $200 top bet × 200 = $40,000 recommended bankroll for 1-to-8 at <1% RoR
Half those amounts at the 2% RoR threshold (more acceptable for trip-only play)

Most amateur counters dramatically under-bankroll their spread. A working AP with $10,000 should be running 1-to-8 at $12.50/unit (top bet $100) — or 1-to-12 at $8/unit, which most casinos don't deal — or flat-betting at $25 with a small bet variation in between. The 1-to-12 ramp at $25/unit on $10,000 produces an RoR around 13%, which is the full-Kelly risk level the working AP should specifically NOT accept.

When 1-to-8 actually wins

Three scenarios where 1-to-8 is the better practical choice despite lower published EV:

Heat-prone property. The 1-to-12 spread gets the player backed off in 3 sessions; the 1-to-8 spread runs for 30 sessions. The forward expected value of the 1-to-8 spread over 30 sessions is much higher than the 1-to-12 over 3, even at the lower per-session EV.
Bankroll under $40,000. The bankroll math says the top bet on 1-to-12 is too big a fraction of the roll; the 1-to-8 ramp's smaller top bet keeps RoR in the working 1% target.
Shorter trip horizons. A 4-hour Saturday session has limited high-count windows. The 1-to-12 ramp's incremental EV over 1-to-8 comes mainly from the TC+5 rounds, which are 4% of hands; over 320 hands you'll see roughly 12 such rounds, and the marginal extra EV from betting $300 instead of $200 on those rounds is about $20 total. The heat exposure across 320 hands of bigger-bet-pattern observation can outweigh that.

When 1-to-16 (or more) makes sense

Two scenarios:

Team play. One player is the counter, another is the big-bet 'gorilla' who only sits down when called over to a hot shoe. The gorilla's spread looks like 1-to-1 from the casino's perspective (every bet is a top bet); the team aggregate spread is effectively unlimited.
One-shot move. A counter playing a property exactly once (e.g., on a trip to a new market where they don't plan to return) can spread as wide as the bankroll allows because there's no future EV to protect. The 1-to-20 spread captures the EV before the backoff lands.

For a regular counter at a regular property, anything above 1-to-12 is heat with diminishing returns. The math chart says 1-to-16 captures 0.90% vs 1-to-12's 0.75% — a 20% EV improvement — but the published heat experience says 1-to-16 gets backed off 5x faster. Forward EV: the 1-to-12 spread wins for most working APs at most properties.

The practical spread guide

If you have $5,000-$10,000 in bankroll: 1-to-4 spread, accept the thin edge, use it as training ground for the count and the cover act. You're not going to grind serious money at this bankroll size; the goal is rep-building and small positive EV.

If you have $10,000-$25,000: 1-to-8 spread, $25-$50 unit. This is the working AP entry zone — meaningful hourly EV with manageable RoR and acceptable heat at most properties.

If you have $25,000-$60,000: 1-to-12 spread, $25-$100 unit (depending on the upper end of your bankroll). The standard professional ramp. Run it carefully at heat-prone properties; run it freely at locals casinos and smaller off-Strip.

If you have $60,000+ and 1-to-12 has been worked at multiple properties: consider rotating multiple casinos at 1-to-12 rather than escalating to 1-to-16 at a single property. The per-property forward EV cap is essentially the same; spreading across properties multiplies it.

Run your bankroll, edge, and target RoR through /calc/risk-of-ruin to back-solve the recommended unit size for any spread, then use the bet-spread visualizer at /advanced/bet-spread to see the per-TC unit allocation that hangs off it. The sibling post at /blog/risk-of-ruin-bankroll-formula covers the bankroll-side math that determines whether your spread is sustainable; /blog/true-count-conversion-explained covers the TC computation that the spread depends on. Drill the ramp itself in the counting trainer at /train/counting and at the table with /counting-live, which shows the recommended bet for the current count in real time so you can spot the moment your spread is misaligned with the math.